GroopM

GroopM is an automated tool for the recovery of population genomes from related metagenomes.[1][2][3] It depends on BamM,[4][5] which is also installed.

UPDATE Aug 2018 The original authors of GroopM seem to have stopped developing it. A new author, Tim Lamberton, has taken over and seems to be doing further development.[6] He has also taken over BamM development.[7] This GroopM is version 2.0.0, which is accessible using the command groopm2. Please use this version, instead.

Installed Versions

GroopM 0.3.5 with BamM 1.7.3[8] and CheckM 1.0.7[9] is installed on Proteus. Because it depends on some obsolete Python modules, it comes with its own installation of Python 2.7.13. Use the modulefile:

groopm/0.3.5

The newer version GroopM 2.0.0 by [10] uses its own Python 2.7.15 and is available with the modulefile:

groopm/2.0.0

This module also loads, among others:

perl/5.20.0
oracle/jdk/1.7.0_current
samtools/1.2
bwa/master
pplacer/1.1.alpha19
prodigal/2.6.3
hmmer/3.1b2

Caveat

The QIIME modules provide different versions of GroopM and BamM because QIIME's scripts depend on specific versions of underlying tools; QIIME may not work properly using non-canonical versions. Undetermined behavior may result if both GroopM and QIIME modules are loaded together.
The same is true of Python -- the groopm module includes its own version of Python 2.7.13 because some of its dependencies are not the latest installed on Proteus.

Using

If you need to use another module which conflicts with the GroopM module, you will need to unload groopm before loading the other module, and vice versa:

module load othermod/1.2.3 dosomething module unload othermod module load groopm groopm something

Please be sure to cite the authors of GroopM and BamM appropriately in any publications. See the official websites for these software for details.

Compiling

The recommended installation method is to use "pip"

Prerequisites

Python 2.7.x
BamM: this one, not that one; which requires
- SAMtools 1.2
- BWA 0.7.12
numpy (for CheckM)
PyTables 2.x
- PyTables 2.x is obsolete and no longer supported. There is an error in its setup.py which causes it to incorrectly check the numpy version number. The fix (diff file) is here: https://gist.github.com/prehensilecode/2eb790476c38299e520ce5ea40896e08